Part 2:Using FMEA, DFR, Test andFailure Analysis in Lean NPD
Overview• Introduction and Definitions• Part 1: Lean Product Development
– Lean vs. Traditional Product Development– Key Elements of Lean NPD
• Customer Defines Value• Front Loaded and Knowledge Based• Eliminate Redesign Waste
– Reliability Requirements• Part 2: Reliability Elements of Lean NPD
– Lean FMEA and DRBFM– Critical Characteristics– DFR and Physics of Failure– Accelerated Testing to Failure– Failure Analysis and Knowledge Capture
Lean FMEA
• Some teams attempt to lean FMEAprocess by creating product family FMEAsbut fail to update FMEA for newapplications or changes
• Instead Lean FMEA Should Focus on NewDesign Features and Changes to BaselineDesign to Assess Associated Risks
Tools to Focus Lean FMEA• Diagramming Tools
– Functional Block Diagram– Boundary Diagram– Parameter Diagram– Process Flow Diagram
• Highlight Changes to Product or Processon the Diagrams
Functional Block Diagram
Annotate Retained and Changed Items & Functions
Boundary Diagram ConstructionSubsystem 1
Subsystem 3
Subsystem 2
FMEA Boundary
Interface-Physical Interface-Info Transfer-Data Transfer-External InputConsider a Functional Block Diagram of the System
With Modules and Interfaces
Parameter Diagram ofProduct, Process, System
Product, Process,or System
NoiseFactors
Response(Output, Function)
Control Factors
Signal(Inputs)
Elements of the P-Diagram
• Controlled by Input Function• Static or Dynamic• May be Variable
• Performance Mean, Std Dev• Customer Requirement
• Forces Beyond Control• Cause Output Variation• Environment Factors
• Functional Design Parameters• Fixed or Adjustable• Fundamental to Design of Function• Reduce Variation
Process Flow Diagram
Annotate Retained and Changed Process Steps
Selecting Process Steps for Analysis
• On Process Flow Map:– Identify Steps Being Modified– Identify New Steps Required for New Product
• Drill Down to Identify Sub-Steps Within theTarget Steps Identified for Analysis
• Complete Lean PFMEA on SelectedProcess Steps
• Integrate with Previous PFMEA onStandard / Unchanged Process Steps
Key Characteristics• Include:
– Product Features– Manufacturing Processes– Assembly Characteristics
• That Significantly Affect:– Product Performance– Form, Fit, Function
• Lean NPD Focuses on the Critical FewCharacteristics the Customer Values
Key Characteristics
DetectPrevent
RPN
DET
OCC
SEV
ActionTaken
Action ResultsResponse &
TargetComplete
Date
RecommendedActions
RPN
Detec
Current
Controls
Occur
PotentialCause(s)/
Mechanism(s)Of Failure
Class
Sev
PotentialEffect(s) of
Failure
PotentialFailureMode
Item /ProcessStep
DetectPrevent
RPN
DET
OCC
SEV
ActionTaken
Action ResultsResponse &
TargetComplete
Date
RecommendedActions
RPN
Detec
CurrentDesign
Controls
Occur
PotentialCause(s)/
Mechanism(s)Of Failure
Class
Sev
PotentialEffect(s) of
Failure
PotentialFailureMode
Item /ProcessStep
Function
DFMEA &PFMEA
Technical Requirements,Tools Identify SpecialProduct Characteristics
Seve
rity Process Steps
Rec
eive
Mat
eria
lM
ater
ial h
andl
ing
Shi
ppin
g D
amag
eC
ompo
nent
Man
ufac
ture
Veh
icle
Ass
embl
y
Primary Drive Manufacturing Process Steps
Op
100
Ste
p 1
PR
E-LO
AD
DO
WE
L P
INS
TO
FIXT
UR
EO
p 10
0 S
tep
2P
RE-
LOA
D J
AC
K S
HA
FTS
EA
L TO
FIX
TUR
EO
p 10
0 S
tep
3P
RE-
LOA
D P
RIM
AR
YH
OU
SIN
G B
US
HIN
G T
OFI
XTU
RE
Op
110
Pre
-load
bea
ring
to f
ixtu
re #
2O
p 12
0P
re-lo
ad m
ain
shaf
t oil
seal
to
man
drel
Op
200
Hou
sing
to f
ixtu
re #
1O
p 21
0O
pera
te p
ress
Op
220
Ret
aini
ng r
ing
to to
p gr
oove
Op
230
Rel
oad
fixtu
re #
1O
p 30
0H
ousi
ng to
fix
ture
#2
Op
310
Ope
rate
pre
ssO
p 32
0R
etai
ning
rin
g to
top
groo
veO
p 33
0M
andr
el t
o m
ain
shaf
t bor
e I.D
.O
P 3
40O
pera
te p
ress
Op
350
Re-
load
fix
ture
#2
and
man
drel
Op
400
Hou
sing
to ta
ble
Op
410
Res
erve
dO
p 42
0C
hain
adj
sub
ass
y to
hou
sing
Op
430
Lubr
icat
e bu
shin
g &
sea
lO
p 44
5M
ove
or s
tage
for
fina
l ass
yO
p 10
O-r
ing
to s
hifte
r tu
beO
p 50
0S
hifte
r tu
be t
o ho
usin
gO
p 51
0C
lam
p to
shi
fter
tube
Op
20A
ssem
ble
shift
er le
ver
Op
520
Wav
e w
ashe
r to
shi
fter
leve
rO
p 53
0S
hifte
r le
ver
to s
hifte
r tu
beO
p 53
5Q
ue fo
r fin
al a
ssy
lineCustomerAssessment
Direction of Improvement
Pote
ntia
l Crit
ical
and
Sig
nific
ant
1.090 TO 1.110 " FACE OF PRIMARY HOUSINGBUSHING TO FACE OF JACK SHAFTSEALDOWEL PINS 0.260 TO 0.270 " TOFACE
JACK SHAFT SEAL AGAINSTSHOULDERBEARING FLUSH TO SNAP RING FACE
SEAL COMPRESSION HEIGHTPRIMARY GASKET SEAL SURFACE FINISH
SERATION DAMAGE
Weighted Importance
Relative Importance
G FH F
H FG F G
H G H H F H HF F Y H H H H H GG
F
5393995
0 81 81 72 0 9 41 0 0 27 54 261
3 0 54 81 3 0 0 0 27 0 45 9 9 0 0 0 0 0 0 0
Special CharacteristicsMatrix
Characteristics Matrix
Requirements Documents•Customer Specification•Regulatory•Dimensions•Appearance
Requirements DocumentDrawingsField History Robustness Tools
•Functional Block Diagram•Boundary Diagram•P-Diagram•Interface Matrix
Identifying Key Characteristics10
987654321
1 2 3 4 5 6 7 8 9 10
Occurrence
Severity
Potential Critical Characteristic
Potential Key Characteristic Possible Annoyance Zone
Special Characteristics Matrix
Seve
rity
Process Steps
Rec
eive
Mat
eria
l
Mat
eria
l han
dlin
g
Shi
ppin
g D
amag
e
Com
pone
nt M
anuf
actu
re
Veh
icle
Ass
embl
y
Primary Drive Manufacturing Process Steps
Op
100
Ste
p 1
PR
E-LO
AD
DO
WE
L P
INS
TO
FIXT
UR
EO
p 10
0 S
tep
2P
RE-
LOA
D J
AC
K S
HA
FTS
EA
L TO
FIX
TUR
EO
p 10
0 S
tep
3P
RE-
LOA
D P
RIM
AR
YH
OU
SIN
G B
US
HIN
G T
OFI
XTU
RE
Op
110
Pre
-load
bea
ring
to f
ixtu
re #
2
Op
120
Pre
-load
mai
n sh
aft o
il se
al t
om
andr
el
Op
200
Hou
sing
to f
ixtu
re #
1O
p 21
0O
pera
te p
ress
Op
220
Ret
aini
ng r
ing
to to
p gr
oove
Op
230
Rel
oad
fixtu
re #
1O
p 30
0H
ousi
ng to
fix
ture
#2
Op
310
Ope
rate
pre
ss
Op
320
Ret
aini
ng r
ing
to to
p gr
oove
Op
330
Man
drel
to
mai
n sh
aft b
ore
I.D.
OP
340
Ope
rate
pre
ssO
p 35
0R
e-lo
ad f
ixtu
re #
2 an
d m
andr
elO
p 40
0H
ousi
ng to
tabl
e
Op
410
Res
erve
d
Op
420
Cha
in a
dj s
ub a
ssy
to h
ousi
ng
Op
430
Lubr
icat
e bu
shin
g &
sea
lO
p 44
5M
ove
or s
tage
for
fina
l ass
yO
p 10
O-r
ing
to s
hifte
r tu
beO
p 50
0S
hifte
r tu
be t
o ho
usin
gO
p 51
0C
lam
p to
shi
fter
tube
Op
20A
ssem
ble
shift
er le
ver
Op
520
Wav
e w
ashe
r to
shi
fter
leve
rO
p 53
0S
hifte
r le
ver
to s
hifte
r tu
beO
p 53
5Q
ue fo
r fin
al a
ssy
line
CustomerAssessment
Direction of Improvement
Pote
ntia
l Crit
ical
and
Sig
nific
ant 1.090 TO 1.110 " FACE OF PRIMARY HOUSING
BUSHING TO FACE OF JACK SHAFT
SEAL
DOWEL PINS 0.260 TO 0.270 " TO
FACE
JACK SHAFT SEAL AGAINSTSHOULDER
BEARING FLUSH TO SNAP RING FACE
SEAL COMPRESSION HEIGHT
PRIMARY GASKET SEAL SURFACE FINISH
SERATION DAMAGE
Weighted Importance
Relative Importance
L H
M H
M H
L H H
M L M M H M M
H H M M M M M M L L
H
5
3
9
3
9
9
5
0 81 81 72 0 9 41 0 0 27 54 261
3 0 54 81 3 0 0 0 27 0 45 9 9 0 0 0 0 0 0 0
Special CharacteristicsMatrix
Process Steps from Flow Chart
Characteristics from Requirements and DFMEAEffect of Step onCharacteristics
Developing the Control Plan• Prioritize Process Risks Identified in the
PFMEA and the Special CharacteristicsMatrix
• Process Flow Diagram• Lessons Learned from Similar Processes• Process Control Data from Related
Processes• Measurements Required for Process Control• SPC Control Limits• Complete the Items in Control Plan Template
Control Plan ItemsMachine, Device, Jig,
Tools for Mfg.For each operation that is described, identify the processing equipment machine, device,
jig, or other tools for manufacturing, as appropriate.
No. Enter a cross reference number from all applicable documents such as, but not limited to,process flow diagram, numbered blue print, FMEAs, and sketches.
Product Features or properties of a part, component or assembly that are described on drawingsor other primary engineering info. Compilation of important product characteristics.
Process Process variables that have a cause and effect relationship with the identified productcharacteristic. Identify those process characteristics for which variation must becontrolled to minimize product variation. There may be more than one processcharacteristic for each product characteristic.
Product/ProcessSpecifications/Tolerance
Specifications/tolerance may be obtained from various engineering documents, such as, butnot limited to, drawings, design reviews, material standard, computer aided design data,manufacturing, and/or assembly requirements.
Evaluation/Measurement/Technique
Identify the measurement system being used, including, gages, fixtures, tools, and/or testequipment required to measure the part/process/manufacturing equipment.
Sample When sampling is required list the corresponding size and frequency.
Control Method Brief description of how the operation will be controlled, including procedure numberswhere applicable. Operations may be controlled by SPC, inspection, attribute data,mistake proofing, sampling plans, and other. If elaborate control procedures are used,reference document by ID name/number.
Reaction Plan Specify the corrective actions necessary to avoid producing nonconforming productsor operating out of control. May also refer to a specific reaction plan number and identifythe person responsible for the action.
Using the Control Plan
• For Critical Characteristics, Use ControlPlan to Identify:– Measurements: How, When, How Often– Controls to Keep Characteristic in Tolerance– Actions if Characteristic Out of Tolerance
• Containment• Corrective Actions
• Robust Design + Controlled Processes =Reliable Products
DRBFM and DRBTRDesign Review Based on Failure Modes & Test Results
Key Elements of Mizenboushi(Reliability Problem Prevention)
GD3 (Good Design, Good Discussion,Good Dissection)
Toyota’s “Creative FMEA” Method
Problem Prevention – GD3
• Good Design = Robust Design– Design for Reliability (DFR)– Design for Six Sigma (DFSS)
• Good Discussion = Eliminate Risk– Apply Design Review Based on Failure
Modes (DRBFM) to identify problems anddevelop countermeasures or corrections
• Good Dissection = Effective Validation– Apply Design Review Based on Test Results
(DRBTR) to Evaluate Effectiveness. Test toFailure & Analysis of Test Failures is Critical
GD3 Problem Resolution
Good DiscussionGood Dissection
Good Design
Development(DiscoverUnknownProblems)
Total Problems to be Solved
Pre-Production
Validation
OBJECTIVE:
Discover& ResolveAll ProblemsBefore Launch
(Robust DesignTo PreventProblems) Unknown
Problems
DiscoveredProblems
DRBFM Approach• Elements from FMEA, FTA, and Design
Review– These tools previously used for management
and control of projects– Toyota developed “creative FMEA” approach– Shift focus to improve perceptiveness and
problem solving• Focus is on finding and preventing
problems – not completing forms andchecklists (which de-motivate participants)
DRBFM Application
DRBFM is a Forum for ThinkingTeamwork and Participation
What has changed? * What did you change? Why?* What surrounding conditions have changed
outside your control?
Concerns about the changes? *Your Concerns? What other concerns?* Draw on expertise and knowledge of past problems
When will concerns appear? *Could concerns become causes of failuresor incidents? Visualize concerns & causes
What effects will there be? *How will causes effect the customer?* Consider effects on the OEM and end user
What preventive measures *What has been done to assure concerns willhave been & should be taken? not actually appear?
*Consider other measures that can be implemented
Source: Bill Haughey, DRBFM, Applied Reliability Symposium, June 2007, March 2008
How is DRBFM Done?
• Preparation for the Design Review• Conducting the DRBFM• Capturing the Inputs• Assigning and Tracking Actions to
Completion
Pre-Work for the DRBFM• Design Engineer or Core Team:
– Functional Diagram, Operating Environment– Changes from Previous Baseline Design– Drawings and Analysis– Failed and Sectioned Parts– Draft DRBFM with Components / Changes, Concerns
with Causes and Factors, Effect on Customer, Designto Eliminate Concerns
• Participants (Functional Experts):– Perceptive mindset, interest in improving product– Past experience and knowledge on similar items
Capturing the Data
Typical DRBFM Session
Source: A Guide to GD3 Activities and DRBFM Technique to Prevent Trouble, Kano & Shimizu, Toyota 2001
Changes and Results Documented
Source: Carl Hanser Verlag, QZ, Munich, 4-2005
Linking DRBFM with FMEA
• DRBFM captures the information neededfor FMEA except scoring
• Scoring columns can be added to fill needfor FMEA if required by customer orstandards
DRBFM to FMEA (1)
FMEA FormAdd Scores
C Potential O D
Item Item Potential Potential S l Cause(s) / c Current eR
.
Function Failure Effect(s) of e a Mechanism(s) c Design tP
.
Mode Failure v s of Failure u Controls eN
.
s r c
DRBFM Form:
DRBFM to FMEA (2)
D
Current e R. Recommended Responsibility Action Results
Design t P. Actions & Target Actions S O D R.
Controls e N. Date Taken e c e P.
c v c t N.
DRBFM Form:
Add Scores Add Scores
FMEA Form:
DRBFM System Integration
Source: SAE Paper 2003-01-2877, Shimizu, Imagawa, Noguchi, Reliability Problem Prevention for Automotive Components
Applying DRBFM in Lean NPD
• Use “Missing Knowledge” Decision Flowfrom Lean QFD as starting point
• These unknowns and known changesfrom current technology or design are thegreatest risks
• Focus DRBFM on these unknown andchanged areas during concept andprototype team reviews & IntegrationPoints
Integrating Product & ProcessDesign
• DFMA – Design for Manufacturing andAssembly
• Integrated Product and Process FMEA /DRBFM
• Concurrent Engineering Team• Visual Management – Decision Flow and
Value Stream Map to Manage Tasks
Impact of Lean Focused FMEA• Allocation of Resources Targeted to
Reduce Highest Risks and Unknowns• Impact Product and Process Design• Drive Test Planning and Analysis to
Resolve Issues and UnderstandUnknowns
• Verify Corrective Action Effectiveness• Critical Characteristics and Process
Measurement / Control
Lean FMEA Summary• Focus FMEA on Changes in Design or
Process• Use Supporting Tools to Narrow Focus:
– Parameter & Boundary Diagrams– Process Flow Charts– DRBFM Techniques– Characteristics Matrix
• Use Lean NPD Tools to IdentifyUnknowns, Apply Resources, AssignTasks
Design for Reliability andRobustness
How do we Design-in Reliability?• Stress Analysis and Test
– Find Product Limits & Understand User Needs– Products fail due to variation or in limit environments
where stress exceeds strength– Stress and strength distributions:
DFR Strategies
Stress-Strength vs. Age
Reducing Stress / StrengthInterference
• Increase strength of the part– Understand operating environment stresses– Select more robust parts or materials– Increase design margin– Supplement deterministic design with probabilistic
tools• Reduce part strength variability
– Understand sources of part variation and deterioration– Controlled production process (SPC)– Protect vulnerable components
• ROBUST DESIGN + CONTROLLEDPROCESSES = RELIABLE PRODUCT
Robust Design Tools
DFSS and DFR Tools: Differences and Commonality
Probabilistic Design
Applied Reliability Engineering, Rousch and Webb, Center for Reliability Engineering, University of Maryland, College Park, MD. Jan 2006
Reliability Based DesignOptimization
Elements of Probabilistic Design
• Understand physics of failure and stressesthat precipitate failure
• Use predictive modeling and acceleratedtest to failure to estimate probability offailure
• Consider variability of applied stressesand variability of product strength
• Eliminate stress-strength interference
Physics of Failure Approach
• Robust Design Considerations• FMEA or DRBFM Methods
– Design Review Based on Failure ModesIntegrates FMEA and Design Review
• Test to Failure and Understand CauseMechanisms
• Failure Analysis Methods
Understand Physics of Failure
• What physical phenomenon in the part is causedby the stresses applied?
• If we understand the root cause, we can improvestrength or reduce variability to prevent ormitigate the failure.
• Most hardware failures can be traced to fourphysical categories / mechanisms:– Wear– Corrosion / Contamination– Mechanical Failure (fatigue, vibration resonance, etc.)– Overstress (electrical or mechanical, transients)
Physics of Failure Tools• Tools Used in Physics of Failure Analysis
– Principal Physics Model– CAD Drawings / Solid Modeling– Finite Element Analysis
• Dynamic Simulation (Transients)• Fatigue Analysis (Cumulative Damage)• Thermal Analysis• Accelerated Testing• Simulation
Design of Experiments (DOE)
• Tool to Evaluate Design Alternatives• Determine Factors and Response• May need Two Phased DOE Approach
– Fractional Design to Find Main Factors– Full Factorial Design to Evaluate Effects and
Interactions on Reduced Set of Factors– Consider Time and Cost
• Analysis of Results and Optimization ofSolution
Iterative DOE Process
Larry Gonzales, Raytheon, Experiment Design for Engineers & Scientists, Applied Reliability Symposium, 2009.
Use Trade-Off Curves to CaptureKnowledge
• Point Data from Analysis and Experiments• Relationship Between Key Parameters• Apply to Support Design Decisions
Trade-Off Curves
Capture Knowledge from Point Solutions
Trade Off Curve Example
Key is Understanding
• Methods Build Knowledge of Alternatives– Physics of Failure– Design of Experiments– Design for Robustness and Reliability
• Enable Better Design Decisions• Eliminate or Reduce Redesign Waste
Reliability Testing and DataAnalysis
Phased Robustness Testing
• Prototype Phase– Accelerated Test to Failure (Well beyond Spec –
HALT, Step Stress, Specific Stresses and FailureModes)
• Design Verification Phase– Quantitative Accelerated Life Test– Selected Qualification Tests
• Production Validation– Demonstrate Corrective Action is Effective– Validate Final Product Made on Production Tools
Robustness Indicator Figure
Factors(Temperature, Vibration,Humidity etc) Analysis & Test Results for
Each Factor on current ornew product
Requirement orSpecification Margin or
Robustness ofDesign Factor
(Can be Created in Excel using Radar Chart)
If RequirementExceeds TestResult or hasSmall Margin,Design is NotRobust
General Approach to AcceleratedLife Test (ALT)
• Understand Failure mechanisms• Understand Operating and Design Limits• Clarify Use Level Stress Application• Conduct Qualitative tests like HALT or step-
stress tests to define product limits and failuremodes
• Conduct Quantitative ALT to extrapolate life atuse level conditions– Times to Failure at Accelerated Stress Levels– Use life-stress relationships and distributions
ALT Plan• Stresses to be Considered• Life-Stress Relationship for Each Stress• Application Use Level for Each Stress• Use Level Failure Criteria / Threshold• Test Duration and Resources Available• Consider Use of DOE to help estimate:
– Stress Factors with Most Effect– Probability of Failure at Specified Use Level– Probability of Failure at Maximum Stress– Interactions to Help Define Life-Stress Relationship
Highly Accelerated Life Test(HALT) – Qualitative ALT
1. Improve Reliability by FindingWeaknesses and CorrectingThem Early.
2. Establish Upper and LowerOperating and Destruct Limits ofEnvironmental Stressors
3. Typically done in Temperature &Vibration Chamber forElectronics & ElectromechanicalProducts
4. Concept can be Applied to OtherStressors (Voltage, Current,Mechanical Loads, etc.)
Quantitative ALT
• Test to Failure at Multiple AcceleratedStress Levels
• Use Analysis to Extrapolate Reliability orLife at Application Use Level Stress
• Can be Used to Demonstrate Ability toMeet Reliability Requirements
Cautions on Acceleration• Understanding product limits helps prevent accelerating
to unrepresentative stresses and failure modes• Time
– Consider Heat Buildup– Effects of Cycling
• Temperature– Material Phase transitions– Non-linear response– High temperature or thermal cycling?
• Power– Protective and limit devices– Transients
• Vibration– Mechanical limits or resonances
• What failure mechanism are we accelerating?
Data Collected• Test Parameters Measured
– Temperature, Power Density, Cycle Rate, Vibration,Humidity, Voltage, etc. applied
– Product Response or Function (monitor during test)• Time to Failure or Run Time (Suspended)
– At least 3 Different Stress Levels– Fit Data Points to Appropriate Distribution
• Product Limits from Step Stress Test• Failure Mode Observations
Data Analysis• Analyze Data and Extrapolate Life at Use Level
Stress• Life-Stress Relationships (predictive models)
– Arrhenius – Temperature– Eyring – Temperature or humidity– Inverse Power Law – Voltage, Power, Mechanical– Multiple Life-Stress Models
• Temperature / Humidity• Temperature / Non-Thermal: Temp / Voltage or Power• General Log Linear: multiple accelerating stresses• Proportional Hazards: multiple covariates• Cumulative Damage – Time varying stress profiles
Data Analysis – Use LevelReliaSoft ALTA 7 - www.ReliaSoft.com
Use Level Probability Weibull
Beta=3.7483; Alpha(0)=-6.0220; Alpha(1)=5776.9341; Alpha(2)=-1.4340; Alpha(3)=0.6242
Time
Unrel
iabilit
y
1000.000 100000.00010000.0000.100
0.500
1.000
5.000
10.000
50.000
90.000
99.000
0.100
Use LevelCB@90% 1-Sided TB
Data 1General Log-LinearWeibull328|10|1F=35 | S=51
Data PointsUse Level LineTop CB-IBottom CB-I
John PaschkewitzWatlow Electric Mfg Co2/4/20083:39:07 PM
Data Analysis - DistributionReliaSoft ALTA 7 - www.ReliaSoft.com
Probability Density Function
Beta=3.7483; Alpha(0)=-6.0220; Alpha(1)=5776.9341; Alpha(2)=-1.4340; Alpha(3)=0.6242
Time
f(t)
0.000 20000.0004000.000 8000.000 12000.000 16000.0000.000
3.000E-4
6.000E-5
1.200E-4
1.800E-4
2.400E-4
Data 1General Log-LinearWeibull328|10|1F=35 | S=51
Pdf Line
John PaschkewitzWatlow Electric Mfg Co2/4/20083:41:01 PM
Life vs. Stress as Trade-OffRe l ia So ft AL TA 7 - w w w .Re l ia So ft.c om
L if e v s S t re s s
Std= 0 .7937 ; K= 7 .1350E-12 ; n= 3 .6556
W ir e W a t t D e n s it y
Lif
e,
Cy
cle
s
50 .000 1000 .000100 .00010 .000
10000 .000
100 .000
1000 .000
L i feCB@ 90% 1 -S ide d TB
Ax c e l i s AL T N iCr W W D L -SI nv e rs e Pow e r L a wL ogno rm a l132 .4F = 5 | S= 1
M ea n L i fe L ineTop CB M e a nBo ttom CB M ea n
157.8S tre s s L e v e l Po in tsM ed ia n Po in tI m pos ed Pd f
186 .7S tre s s L e v e l Po in tsM ed ia n Po in tI m pos ed Pd f
281 .6S tre s s L e v e l Po in tsM ed ia n Po in tI m pos ed Pd f
J ohn Pa s c hk ew itzW a tlow E le c tric M fg Co10 / 8 / 20093 :06 :05 PM
1 3 2 1 8 7 2 8 2 5 0 0 7 0 0
R e g i o n o fU n a c c e p ta b l eW a tt D e n s i ty
, w s i
R e g i o n o fA c c e p ta b l eW a tt D e n s i ty
A LT C y c les f rom 250 t o 400 C
A LT C y c les f rom 250 t o 430 C
A LT C y c les f rom 250 t o 500 C
0 . 168" d ia s hea t h. 021 " d ia P C N iC rw ire c y c led us ing
c on t ro lled du t y c y c leram p ov e r 78m inu t e c y c le
Production Validation Test
• Repeat selected Qualification Tests onany Changes in Product or Process
• Test Samples made on ProductionProcesses
• On-Going Reliability Test (ORT)– HASS – Highly Accelerated Stress Screening– HASA – Highly Accelerated Stress Audit– Periodic HALT Re-Test on Production Units
Failure Analysis / FRACAS
Failure Analysis Tools• Basic
– Recovery of Failed Samples– Electrical Test, Microscopy, Digital Photography
• Non-Destructive Methods– X-Ray (Real Time Digital is Particularly Helpful)
• Disassembly / De-capsulation– Tools or Chemicals to remove layers
• Scanning Electron Microscopy & EDS– Defects, Corrosion, Material Failure
• Acoustic Microscopy / Imaging (Voids / Defects)• Some Internal, Others at Outside Labs
NDT – Electrical Characteristics
Curve Tracer – Showing Good and Failed Part Response
Real Time Digital X-Ray
Real Time X-Ray Example
Digital X-Ray Examples
SEM / EDS
Example of SEM / EDS Analysis
Pt: Rh: Al87:10 : 3 Pt : Rh : Si
82.7:10.3:6.7
PCB Failure Analysis Methods
Thomas Paquette, Insight Analytical Labs, Test & Measurement World, August 2006.
Failure Analysis Summary
• Progressive Use of Tools from NDT toDissection and Cross-Section Exam
• Objective is to Find Physical Evidence ofFailure Mechanism
• Document with Photos and Analysis toCapture Knowledge Gained
• Update FMEA or DRBFM with FA Findings
FRACASFailure Reporting, Analysis & Corrective Action System
• Build Knowledge Base• Process and Tool
– Reliability & Quality of Product, Service, Process orSoftware is Tracked, Measured, and Improved
– Applies to Entire Product Life Cycle– Consistently Ranked Among the Most Important
Reliability Tasks– Closed Loop: Ability to Feed Root Cause & Corrective
Action Information Back Into Design Process forFurther Improvement
Hierarchy of Failure Causes
Design Related Failures
Manufacturing & QualityRelated Failures
Repair Induced
CustomerInduced
Capture Failures from Verification Test to Field Operation
Capturing FA Knowledge• Capture Failure Analysis Results in
Searchable Tool– Commercial Data Base Tool– A3 Format Documents with Keywords– SharePoint (Microsoft)
• Key is Ability to Retrieve Knowledge withMinimal Search Effort
• Lean NPD is Knowledge Based – Key isContinually Adding to AccessibleKnowledge
Summary• Features of Lean NPD for Reliability
– Front End Focus to Gain Knowledge• Basis for Better Design Decisions
– Design for Robustness, Reliability– Understand Physics of Failure
• Testing to Learn and Verify– Test to Failure and Understand Causes
• Knowledge Capture for Future Re-use• Develop & Control Critical Characteristics
References• King, John P. and Jewett, William S.;
Robustness Development and ReliabilityGrowth; Prentice-Hall, Boston, 2010.
• SAE International, J1211, April 2009, Handbookfor Robustness Validation of AutomotiveElectrical / Electronic Modules.
• Robustness Validation Manual, ZVEI, January2010; www.zvei.org/RobustnessValidation
• Jusko, Jill; New Models for ProductDevelopment, Industry Week, April 21, 2010.
• Morgan, James and Liker, Jeffrey; The ToyotaProduct Development System, ProductivityPress, New York, 2006.
References• Sarakakis, Georgios; Fundamentals of Life Data
Analysis: Concepts and Applications, Tutorial, 2010Applied Reliability Symposium Proceedings, June 17,2010.
• Wiggins, Brian, “A Simpler Look at ProductDevelopment”, Product Design & Development, October8, 2010.
• Mascatelli, R., The Lean Product DevelopmentGuidebook, 2007.
• Soderborg, Dr. Nathan, Lean Product Development,WCBF DFSS Conference, Feb. 2008
• Using DOE Results in Design of ALT, Reliability Edge,Vol 10, Issue 2, pp. 1-7.
References• SAE 2003-01-2877, Reliability Problem Prevention
Method for Automotive Components, H. Shimizu, T.Imagawa, H. Noguchi.
• A Guide to GD3 Activities and DRBFM Technique toPrevent Trouble, S. Kano, H. Shimizu, Toyota, 2001.
• Lean Product Development, Eric Rebentisch, Oct 5,2005, MIT Open Courseware,http://ocw.mit.edu/courses/aeronautics-and-astronautics/16-852j-integrating-the-lean-enterprise-fall-2005/lecture-notes/8_rebentisc_leng.pdf
• Bill Haughey, DRBFM, Applied Reliability Symposium,June 2007, March 2008
• Lean FMEA Training, Quality Associates International, atwww.quality-one.com
• AIAG, FMEA, 4th Edition, June 2008, pp.135-138.
Feedback / Follow-up
• Please provide your feedback on this webbased short course: e-mail [email protected] or to ASQ RD
• One day seminar on this topic availablethrough:http://www.hobbsengr.com/Accelerated_Reliability_Seminar_Schedule.htm– Apr 13, 2011 in Chicago, IL– May 2, 2011 in Minneapolis, MN
Top Related